Originally posted by JinDesu Lightroom/PS output - Resize to 16MP
That will make the noise equivalent.
Alternatively, printing at any equal size will show equal or better noise for the k-3.
I tried resizing (to 16 MP) a DNG from the IR site for the K-3 at ISO 51200, and compared the result to a DNG for the K-5IIs at ISO 51200, after first using the same noise reduction parameters in Lightroom. The colour noise from the K-3 is much worse even after resizing.
I printed DNGs from the IR site for the K-3 and K-5IIs at ISO 51200, at A3+. (Without resizing). The noise from the K-3 was much worse at plausible Lightroom parameters. I had to raise the noise reduction for the K-3 to amounts that I consider unreasonable to get a match.
(I don't know if I can recover the degradation caused by the excessive noise reduction for the K-3. Perhaps with practice I can get back to the level of the K-5IIs).
Can you provide evidence that what you say works in practice with these two cameras? Not just in theory?